The paradox of preference for unreliable reinforcement: The role of context and conditioned reinforcement.

The authors looked at old pigeon data. The birds picked schedules that paid off only half the time.

They asked why any creature would choose the worse deal. Their answer: signals during the wait act like extra candy.

Pigeons leaned toward the unreliable side if two things happened. A light or sound came on during the delay. That signal itself became reinforcing.

The setup made bad odds feel good. The birds worked harder for 50% pay than for sure pay.

Geckeler et al. (2000) later tested kids with ID. Same trick worked. The children also chased the shaky schedule when a fun stimulus marked the wait.

Perez et al. (2015) pushed it further. They used signals that sometimes meant no food at all. The pigeons still picked that side.

McDevitt et al. (2016) reviewed many bird studies. They concluded that "good-news" cues are so powerful they can wreck optimal choice.

Together these papers show the effect jumps species. From pigeons to children, signals can override the real payoff rate.

Your client may stick with problem behavior that only pays sometimes. Check what cues sit between the response and the reinforcer. A fun song, a flashing toy, or even your own smile can act as conditioned reinforcement. Strip those signals or move them to the better response. You might shift preference toward the reliable, adaptive choice.